gmane.mail.mutt.user

Re: Mutt eating the tab character in headers [some really long text I added upon message creation to see how the wrapping is handled]

Derek Martin — 2012-05-25T18:02:10

It's probably worth noting that this is less of an issue for most
modern clients, because they are GUI apps that use text boxes to
input/display the relevant headers as you're composing (and thus do
not require any fancy machinations to display the headers neatly), and
(typically) just dump the entire message in a read-only text box or
similar control when reading.

This issue originally arose in Mutt because it formats the message for
editing, it wraps long headers with tab characters, and sent the
message out that way.  But some mail clients don't correctly handle a
tab character in displayed headers.

It seems your specific problem (adding custom headers outside the
editor) is a case that wasn't thought of / tested when the original
bug was fixed...  Or, it was (I myself made some mention of replacing
tabs with spaces late in that bug) and deemed to be acceptable loss.
It is odd though, because all (or nearly all) of the other long
headers in your message are apparently wrapped with tab characters.

Re: Mutt eating the tab character in headers

Aaron Toponce — 2012-05-25T16:44:02

Paying closer attention to some other headers, it appears that some
software uses the <tab>, some use a single <space>, and some use 8
<spaces>. Also, RFC 2822 only mentions what happens on the unfold, not how
a fold itself should be created, although I can understand the argument on
why a <tab> would be "a bad idea".

At any event, it doesn't appear that there is any standard on the matter,
so I'll adjust to Mutt's default behavior, now that I understand it a bit
better.

Thanks,

Re: Mutt eating the tab character in headers

Derek Martin — 2012-05-25T15:50:07

There's a lot to wade through, but the answer is here:

  http://dev.mutt.org/trac/ticket/2995

It turns out this is kind of hard to get exactly right.

Re: Mutt eating the tab character in headers

Chris Burdess — 2012-05-25T15:25:53

It is totally equivalent to the space character in that situation (which
is called "LWSP-char" in the specification). That is, they are
interchangeable with no syntactic or semantic loss.

I guess that space has a very minor advantage over tab in this role,
which is that in a text editor that may convert tabs to a series of
spaces, additional spaces would be considered to be part of the field
body contents and not the LWSP-char. So it's safer if you ever intend to
edit the whole RFC822 message in such an editor. Otherwise, it's much of
a muchness.

Re: Mutt eating the tab character in headers

Aaron Toponce — 2012-05-25T15:05:34

I certainly agree. I guess I'm more or less curious why <tab> is being
replaced by <space>- what the logical argument is. Is the <tab> character
not a valid character in mail header fields as defined by the RFCs?

Re: Mutt eating the tab character in headers

Chris Burdess — 2012-05-25T14:24:38

I would think that it doesn't actually matter whether mutt does this or
not, since any intervening MTAs are free to do this as they want. As
long as the result is valid RFC822, header whitespace may be changed or
whatever. Only the values are significant.

Mutt eating the tab character in headers

Aaron Toponce — 2012-05-25T12:46:06

I have an "easter egg", if you will, in the header of my mail. I have two
headers that I am adding: "Crypto-Challenge" and "Crypto-Hint". It's all
for fun and games.

However, in my muttrc(5), I am wrapping each line (It's rather lengthy
otherwise) and preceding the newline with a <tab> character. It shows up
fine when I compose my mail, but when saving, Mutt converts the <tab> to a
<space> (as you can see in the headers of this mail).

For those who don't want to examine the mail headers, here are two
pastebins of what I'm talking about. This: http://ae7.st/p/42a turns into
this: http://ae7.st/p/5f8.

I guess it doesn't really matter, except when viewing headers, other line
wrapping is preceded by a <tab>, and Mutt displays it accordingly. So, why
is Mutt converting my <tab> ta a <space>?

Thanks,

--
. o .   o . o   . . o   o . .   . o .
. . o   . o o   o . o   . o o   . . o
o o o   . o .   . o o   o o .   o o o

Unable to Post to Google Groups: mutt Problem, or Groups Problem?

Koralatov — 2012-05-24T11:20:05

Hi everyone,

I don't seem to be able to post to Google Groups.  I receive the emails
just fine, but when I reply, they send fine, I don't get them bounced
back, but they never appear.  The Groups in question are unmoderated, so I
shouldn't be getting caught in a moderation queue.
        I use a send-hook to send them from the correct address:

send-hook "~t < at >mutt.org"             "my_hdr From: Koralatov <usenet< at >koralatov.com>"
send-hook "~t < at >mutt.org"             "set signature=~/.mutt/sigs/mutt"
send-hook "~t < at >googlegroups.com"     "my_hdr From: Koralatov <lists< at >koralatov.com>"
send-hook "~t < at >googlegroups.com"     "my_hdr Reply-To: lists< at >koralatov.com"

It works just fine for the mutt list, but I'm wondering if maybe it's
what's causing the issue with Google Groups.
        Anyone else having the same problem, or encountered it and managed
to fix it?  The "Reply-To" header was added because I read that an
incorrect Reply-To may result in your message being rejected.

Re: Compiling mutt without mail group

Cameron Simpson — 2012-05-18T23:05:55

| My grateful thanks for the comments and advice.
| Once I've packaged mutt for microcore (with mail group) and reboot
| into stock system, I can send an email via mutt without mail group via
| smtp fine. Does mutt need the mail group for general email, maintenance,
| or interaction with other software such as sendmail ?

Please bottom post. Thanks.

Mutt needs setgid-mail on some (many?) systems to operate on the system
spool file (/var/spool/mail/<user>, typically). It need no special
privileges to operate on the user's files in their home directory.

Re: mutt, html and chrome

Paul Tansom — 2012-05-18T15:49:09

** Luis Mochan <mochan< at >fis.unam.mx> [2012-05-15 21:24]:
** end quote [Luis Mochan]

A bit late to the party here, but I get the same with Chromium on Ubuntu. I
never have the same problem with Firefox though, and since I always get fed up
with Chromium quite quickly and go back to Firefox it doesn't worry me. It will
be interesting to have a play with the suggestions though. I've sort of assumed
it is down to the way Chrome handles it tabs and sub-processes compared to
Firefox - either passing off to another sub-process causing Mutt to delete the
file, or the first process trying to access the file too quickly. I've always
found the interface annoying on Chrome and, unlike many people I know, found it
slow and a bit of a memory hog compared to Firefox! I'm running more modern
hardware now, so perhaps time to see if that has switched the tables!

Re: mutt, html and chrome

Luis Mochan — 2012-05-16T15:14:06

Dear Marco,
Of course, simplicity is usually best! Thanks!
Luis

Re: mutt, html and chrome

Marco Giusti — 2012-05-16T13:34:57

This seems more like an hack but it works most of the times:

text/html; /usr/bin/sensible-browser '%s' & sleep 2; ...
text/html; /usr/bin/iceweasel '%s' & sleep 2; ...
...


I don't know

m.

Re: mutt, html and chrome

Luis Mochan — 2012-05-16T02:34:17

I made a small perl script (attached) that replicates the behavior of
mutt_netscape by Gary A. Johnson <garyjohn< at >spk.agilent.com> and which
seems to solve my problem of disappearing temporal files before the
browser reads them. 
Thanks for your kind help!
Luis




On Tue, May 15, 2012 at 07:30:06PM -0500, Luis Mochan wrote:

Re: mutt, html and chrome

Luis Mochan — 2012-05-16T00:30:06

Thanks Michael,
I'll try that when I get to my office. I found an alternative solution at 
   http://www.spocom.com/users/gjohnson/mutt/#html 
by running a simple script that saves the file and then call's the browser.
Best regards,
Luis


On Wed, May 16, 2012 at 12:27:19AM +0200, Michael Ludwig wrote:

Re: mutt, html and chrome

Michael Ludwig — 2012-05-15T22:27:19

Luis Mochan schrieb am 15.05.2012 um 15:22 (-0500):


IIRC there was an issue where the file would get deleted before the
browser would get its hands on it. There's a set of Python scripts
that fixes this problem:

https://bitbucket.org/blacktrash/muttils

Relevant Mutt configuration:

# call viewhtmlmsg from macro
macro index,pager <F7> "\
<enter-command> set my_wait_key=\$wait_key wait_key=no<enter>\
<pipe-message>viewhtmlmsg -k0<enter>\
<enter-command> set wait_key=\$my_wait_key &my_wait_key<enter>\
" "view HTML in browser"

macro index,pager <F8> "\
<enter-command> set my_wait_key=\$wait_key wait_key=no<enter>\
<pipe-message>viewhtmlmsg -k0 -s<enter>\
<enter-command> set wait_key=\$my_wait_key &my_wait_key<enter>\
" "view HTML (safe) in browser"

Michael

Re: mutt, html and chrome

Luis Mochan — 2012-05-15T20:50:56

I'm using the system's defaults since some long time ago. I guess the
relevant lines in my /etc/mailcap are:

  text/html; /usr/bin/sensible-browser '%s'; description=HTML Text; nametemplate=%s.html
  text/html; /usr/bin/iceweasel '%s'; description=HTML Text; test=test -n "$DISPLAY";  nametemplate=%s.html
  text/html; /usr/bin/w3m -T text/html '%s'; needsterminal; description=HTML Text; nametemplate=%s.html
  text/html; /usr/bin/lynx -force_html '%s'; needsterminal; description=HTML Text; nametemplate=%s.html
  text/html; /usr/bin/w3m -dump -T text/html '%s'; copiousoutput; description=HTML Text; nametemplate=%s.html
  text/html; /usr/bin/html2text '%s'; copiousoutput; description=HTML Text
  text/html; /usr/bin/lynx -dump -force_html '%s'; copiousoutput; description=HTML Text; nametemplate=%s.html

Years ago I used to have 
  text/html; w3m %s; needsterminal; description=HTML Text; nametemplate=%s.html
  text/html; w3m -dump %s; copiousoutput; description=HTML Text; nametemplate=%s.html
in my own .mailcap,

Re: mutt, html and chrome

Matthias Apitz — 2012-05-15T20:30:50

El día Tuesday, May 15, 2012 a las 03:22:20PM -0500, Luis Mochan escribió:


Looks more like mutt is handing over the URI file:///tmp/mutt.html to
your browser and before this has time to fetch the file, mutt has
deleted it again; depends on your configuration in .mailcap

HIH

matthias

mutt, html and chrome

Luis Mochan — 2012-05-15T20:22:20

Very recently I started having a curious situation when I open the
html version of messages with mutt. My default browser is
'chrome'. Typically, in mutt I press 'v', I find the html message from
the list, I press return and the message is displayed in a new window
in my running 'chrome'. Nevertheless, I sometimes get the error
message 'No webpage found for the web address:
file:///tmp/mutt.html'. If I insist enough times, the message
eventually is successfully displayed in my browser. My guess is that
there is some kind of race condition which becomes apparent only when
my computer is busy, as if the browser tries to read the temporal file before
mutt finishes writing and closing it. My system is debian/stable and I
am using the mutt package from the distribution. Any idea/suggestion will be
appreciated. 

Best regards,
Luis

pd. The output of mutt -v is:

Mutt 1.5.20 (2009-06-14)
Copyright (C) 1996-2009 Michael R. Elkins and others.
Mutt comes with ABSOLUTELY NO WARRANTY; for details type `mutt -vv'.
Mutt

Re: Anyone using Mutt on a Mac?

SK — 2012-05-15T04:35:45

If you are not satisfied with or have issue with the default set of
patches included in homebrew formula, you could try
https://github.com/skn/homebrew/tree/master/mutt

Regards,
skn

Re: Anyone using Mutt on a Mac?

Cameron Simpson — 2012-05-14T23:36:08

| now, mutt on iOS would be nice... ;-)

I've tried to use mutt on an iPhone (ssh client to home server). It is
very painful. And the keyboard takes up a LOT of display space.

A retina display iPad may be a lot better.

Re: Anyone using Mutt on a Mac?

Joost Kremers — 2012-05-14T17:08:56

i second that.

i used macports for quite a while but at some point i got fed up with the
fact that it insists on having a completely separate tool chain. macports
installs its own gcc and glibc, for example. makes for looooong upgrades.

but mutt runs fine with both macports and homebrew. i use the same muttrc
on my linux (home) and mac (work) machines, save for some settings
regarding archiving of mail, and it all works without modifications.

now, mutt on iOS would be nice... ;-)